Cancer epidemiology has been identified as a critical area for further progress in the control of cancer. The value of classical epidemiology for insights into etiology and for preventive oncology is well recognized. The application of biostatistical and epidemiological techniques to patients care problems, called clinical epidemiology, has also led to very interesting and useful results. The proposed research program has been design to demonstrate methodological breadth in these two fields. One project is a classical epidemiological case control study of asbestos-containing cigarettes as a risk factor for the development of malignant mesothelioma. This proposal is abased on a recently completed cohort study showing very high mortality in men who made the filters for the cigarettes, which contained crocidolite asbestos. Two projects apply clinical epidemiology techniques, collectively known as risk stratification, to separate clinical problems in cancer. Using these techniques, patients are separated into subgroups at distinctly different risks of an important outcome or disease state. This requires a systematic approach, from hypothesizing potential risk factors, collecting risk factor and outcome data in one patient population, modelling mathematical relationships between risk factors and outcomes using multivariate techniques, and testing these models in additional populations. Only after this process is completed can altered treatment strategies using this verified prognostic information be devised and tested. The current study applies these techniques to cancer patient populations: those with fever and neutropenia and those with suspected spinal cord compression. The proposed studies of patients with fever and neutropenia build on a retrospective study which identified powerfully-predictive prospective risk factors for medical emergencies. The studies of patients with suspected spinal cord compression are more exploratory, based upon the published literature, possibly relevant results of the study of fever and neutropenia and theoretical factors. At the same time, the principal investigator will obtain degree the Master of Science Degree in Clinical Biostatistics at the Harvard School of Public Health, which was specifically designed to give quantitative research skills to clinical researchers. In summary, the proposed research program will promote the training of clinical investigator through an integrated program of classroom training and clinical research. The research plan includes three separate investigations, diverse in their subject matter, which will apply the quantitative material learned in the classroom to significant patient care problems. The end result of this program will be a well-trained clinician, prepared for an independent academic career in clinical cancer research.